Insanity is doing the same thing over and over again, but expecting different results. This saying is attributed to the famous socialist (and theoretical physicist) Albert Einstein but actually traces back to novelist Rita Mae Brown. I have been thinking about this saying a lot lately. The reason is W Social, which is Europe’s newest self-proclaimed attempt at building a competitor to the dominating US- and China-based social media platforms. In many ways, W Social is a poster child for the push of European companies towards digital sovereignty. And in many ways it demonstrates what is so problematic about digital sovereignty. Aline and I have already written about how digital sovereignty is a breeding ground for nationalism and how the idea of “European alternatives” is reproducing the flawed idea of “good capitalism“: The mistaken idea that European companies are immune to the very dynamics that made Big Tech into what it is. W Social is nothing new regarding both points but it illustrates them so vividly that a closer look is warranted even after the attention to W Social has – for now – mostly dried up.
What is W Social?
W Social describes itself as a social network for “verified humans”. A network that looks similar to other popular microblogging platforms (e.g. Threads, Mastodon, X/Twitter or Bluesky). A key marketing element is the required government ID that is needed to set up an account. With this, W Social promises to fight digital violence, hate speech and misinformation because “verified humans” supposedly behave properly online (we will come to what “proper” seems to mean for the owners of W Social). W Social is also advertising their service as being governed by European data protection law and all data being hosted on European infrastructure. Finally, the marketing focuses a lot on being a trustworthy and profitable advertising space for European media and news outlets.
Now let’s look at the organisational and technological facts. What is known about this side of W Social is owed in considerable parts to author and activist Elena Rossini who has done a lot of investigative writing about W Socials. I highly recommend reading her articles. To summarise them: W Social is a platform run by a Swedish startup, backed by venture capital and headed by CEO (and former Chief Privacy Officer of ebay) Anna Zeiter. The people behind W Social are (and I am quoting the FAQ here) “experienced entrepreneurs and investors from the media, technology, and AI sectors”. Technically, it builds on the “Authenticated Transfer Protocol” (AT Protocol, or: atproto). Atproto was developed by former Twitter employees and was first used by Bluesky Social PBC, a US-based for-profit corporation that, among other things, operates Bluesky, a social media platform with (a shrinking number of) 45 millions users worldwide. Atproto enables some interoperability which would – in theory – allow users on W Social to interact with users on Bluesky and other platforms that use elements of atproto (like Eurosky and Blacksky). To verify its users, W Social requires a separate app called W Identity but promises to delete all personal data after the verification process (more on that promise later). When it comes to the benefits for media and news outlets, it is not clear to me what the pitch is exactly. Anna Zeiter has referred to 50 billion euros being spent on social media ads in the EU every year and that W Social hopes to keep that money in Europe. But there are few details on what W Socials is actually planning here. Regardless, some high profile media outlets like the German investigative journalist group correctiv.org already joined W Social (and were also featured on the main page for having joined recently).
W Social soft launched in June of 2026. You still can’t register but have to apply via a waitlist. It currently has about 24.000 accounts but very few are actually active. Contrary to some early reporting the European Union does not appear to have invested public funds into W Social. Yet, there is notable public support from several European officials. Ursula von der Leyen opened up an account early. This could be explained by the platform aligning so evidently with current trends in EU policy like calls for digital sovereignty or pushes for online age verification.
Understanding W Social
There are three aspects I want to focus on specifically: Firstly, I will explain how much of a full social media platform WSocial is (not) right now. Secondly, I will look at the interplay between privacy, GDPR and user verification. Thirdly, I will examine the general promise of trustworthiness. I will then summarise how in every one of these areas W Social misrepresents what it actually does or openly announces intentions that are highly objectionable to begin with.
W Social is not (yet) a social network
As of now, W Social seems to be less a “social media platform made in Europe” as its marketing promises but more like a service that hosts some user data while remaining dependent on the Bluesky infrastructure for much of what makes a platform. To explain this, I need to summarise what atproto is. It contains several key elements:
- Individual identities with follows, posts and media (called “Personal Data Server”, PDS, in atproto vocabulary),
- servers that crawl and index the PDS and map their interactions (called “Relays”),
- servers that publish changes over time in the network (Firehose) and
- software that turns the firehose data into something that can be consumed by end-user-clients, including moderation tools (AppViews).
All these elements are open source and can therefore – in theory – be set up and run by anyone, including W Social. Yet, it took a while for W Social to actually acknowledge that it uses atproto. After initially taking down its public repository, W Social now at least states that it uses the AT Protocol. But, as Aline already explained, using open source software in itself is not saying a lot about whether a specific software serves the public good. It is very telling, however, that W Social did not even acknowledge that it used open code initially. But there is more here than just a lack of transparency. W Social is also using misleading language when describing how much of a platform it is. On the front page, it speaks of being a platform hosted in Europe, but in the fine print (its rather shallow FAQ to be precise), W Social is only speaking of operating “European owned infrastructure for key parts of its service, including Personal Data Server (PDS) infrastructure, moderation services and other platform components” while hinting at an appview in development and referring to “components originally developed by Bluesky” for other functions.
On a technical level, it is understandable that W Social starts with hosting PDS. PDS are probably the most decentralised aspect of atproto. It requires a fairly small amount of resources and skill to run a PDS (for yourself or as a service for others). Other alternative atproto projects like Blacksky or Eurosky started the same way and offered tools to migrate Bluesky accounts to their infrastructure. The difference between W Social and similar undertakings is that others were and are considerably more specific about what parts of atproto they run, while W Social is suggesting that it is a complete platform when this is not the case.
Living up to such a claim in the atproto universe would require much more than running PDS. It would require to also index and map all interactions (relay) and collect the changes into a stream (firehose) to be processed into a usable social graph (appview). Running these parts of an atproto platform is much more resource-intensive than merely hosting PDS. Relay, for example, could be compared to search engines. Both require substantial processing resources and are only feasible for very large organisations. Relays are the elements that got many to question whether atproto can be called decentralised at all. Things seem to have improved since and it seems to be more feasible now to also run your own relay. While this may decentralise the atproto ecosystem in the future, this is not yet the case. Public databases show only 16 atproto relays, several operated by the same entities. It was only recently that Eurosky – after the FreeYourFeeds campaign with support from prominent people like Cory Doctorow, Mark Ruffalo or Alexandra Geese – was able to gather funding and set up their own relay. Blacksky is another project that achieved this. Both Eurosky and Blacksky are also about to or already have published their own appview. Both can be considered independent platforms by now. It is certainly possible for W Social to achieve this as well and there is an appview in the works apparently. But as of now: It has not. Right now, it still relies on infrastructure run by Bluesky which makes W Social sound more like sovereignty washing.
W Social could have gone another route. It could have gone with ActivityPub, the protocol behind the Fediverse, including its most prominent application Mastodon. In the Fediverse every server (called “instance”) includes all aspects of the platform. Every Fediverse server hosts the accounts, federates their interactions to the network and turns this data into something that apps can display. Mastodon has around 2 million monthly active users across 9000 independent servers (caveat: many users are on a few large instances). It would have been great to see even more resources put into the Fediverse, especially after Mozilla shut down their instance (while simultaneously being back on X/Twitter) and the European Commission nearly killed its instance as well. W Social, however, decided to use atproto. The motives behind this decision aren’t public to my knowledge, but from a commercial perspective, atproto makes more sense. Atproto may in theory allow for interoperability, but the protocol still favours a network of few big entities. Using atproto creates – by design – a much more centralisation-prone platform. This is especially visible in its dependency on few relays, while Mastodon consists of thousands of servers that are completely independent from one another and are completely unaffected by other servers going offline. Mastodon instances, while still being compatible with being run for profit, are not as easy to monetise because users can easily move their accounts to thousands of alternative servers. Atproto simply leads to more user lock-in. It therefore seems to be the more suitable choice from a commercial, capitalist perspective. And W Social has made it very clear what their platform is about: “Yes we have a business model […] we actually want to make money.” said CEO Zeiter in a May 2026 presentation that later disappeared from Youtube, but was thankfully analysed by Elena in one of her blog posts.
Privacy and Verification
Another marketing focus of W Social is its supposedly high privacy standard. What this means is not clear. W Social does not advertise specific procedures that go beyond standard measures. The marketing seems to promise compliance with European data protection law (especially its General Data Protection Regulation, GDPR) and W Social may expect to be applauded for not outright disregarding the legal obligations set up by GDPR. This admittedly low standard is not reassuring due to several reasons: The GDPR may have a reputation for being strict but it actually is very compatible with commercial data exploitation and did little to reign in Big Tech corporations. The GDPR is also under attack and the EU Commission is working on lowering its standard in the hope of fostering competitiveness of the European data economy. There is nothing that makes me expect W Social to become a defender of GDPR if it were to become a big European social-media platform.
There is also the issue of user-verification that leaves some questions for me. W Social promises in its FAQ that “ID data is used only during the identification process, after that the data will be deleted”. The terms of service further specify that “every account must be linked to a verified human identity, but we do not collect, retain, or access personally identifiable information sufficient to determine the actual identity of our users.” The privacy notice of W Identity (the verification provider) goes into more detail and clarifies that the direct ID data is being deleted after the process but “verified identity attributes are stored in the W Identity app” and can be stored on the W Identity server by user demand. These verified attributes would allow you to submit proof of being verified later without submitting ID data again (a token in technical language). W Social allows users to use their legal name or a pseudonym. Famously, Yariv Adan, advisor to W Social and AI entrepreneur, uses the pseudonym “Home Simpson” on W Social. But what would happen if a person with the actual name of Homer Simpson would like to use that name? What if Disney (the company that now owns 20th Century Fox and their property, the TV show “The Simpsons”) wants to create a marketing account for the character of “Homer Simpson” and a user needs to prove that he is not using a pseudonym but simply has the same name as the fictional character? What if the above-mentioned Yariv Adan had decided to use his real name as a username as well and another Yariv Adan claimed to be impersonated by him? In these situations one of the parties may be required to not only prove that ID was used to create the account but that the ID actually justifies the use of that specific account name. This is why most ID verification processes usually have to store the original ID for much longer than just for the verification process to finish. Verification is not a one-off action. The “link” that the terms of service refer to needs to stay intact for much longer. Otherwise ID data could be used to create countless or fake accounts which would defeat the purpose of a platform with only “verified humans”.
This brings us to the issue of “verified humans” in general. Verifying users and privacy are simply not compatible. If verification (even if restricted to certain attributes like age) is taken seriously there is no more anonymity online, as the Electronic Frontier Foundation (EFF) and many others keep repeating. The loss of privacy is not even the main issue though. Privacy has become a cyberlibertarian dogma. The EFF has also used it to challenge taxing of crypto gambling and the contemporary understanding of privacy has been weaponised against solidarity, democracy and the public good. The main issue with “verified humans” is what Thomas Dekeyser calls “ontological policing” in his book “Techno-Negative”. As Aline wrote in her review of the book: Ontological policing means drawing a line between who is considered to be, to exist and who is not. Access to government ID is a tool to control people, their lives, their labour and their freedom. Not everyone has access to government ID. In many countries ID is conditional on being employed for migrants which makes these people dependent on (often highly exploitative) employment and some countries like Germany punish political activism (e.g. posts in solidarity with Palestinian people) by stripping people of citizenship. Building a social-media platform on this foundation creates a digital world with a thinking of supremacy built in. And it shows.
Trust your fascists
Everything explained so far serves as the foundation for the key promise of W Social. The platform wants you to “trust your feed” and promises to “prioritise authenticity, safety, and free speech” (FAQ). In practice, this translated into what can only be described as a general atmosphere of European exceptionalism, white supremacy and even outright fascist sympathies. When a platform says to prioritise “free speech”, this usually means that Nazis are allowed. And this is exactly what CEO Anna Zeiter said when clarifying that Alice Weidel, head of the German fascist party AfD, is welcome on W Social. Anna Zeiter later also pushed anti-migration content posted by a right-wing outlet headed by right-wing extremist Martin Sellner. W Social is not (yet) Truth Social, but its leading personnel makes it very clear what the corridor of acceptable positions will look like. All of this is shrouded in progressive vocabulary and European values which brings us to why W Social fits so well into the dominant narrative of “digital sovereignty”.
W Social is digital sovereignty at its worst
To come back to the Albert Einstein quote from the beginning, I actually don’t believe that the people behind W Social are insane for doing what their competitors do, but expecting different results. Because they do not expect different results. W Social is not designed to be different. It is a commercial social-media platform, designed to scale, create profit for investors and serve the dominant political powers in its region of the world. It uses narratives like “digital sovereignty” to appeal to current political leadership (while underdelivering on its technical aspects), promotes a segregation between people that are considered human and those that are not and packages all of it in a language of European and white supremacy. W Social gives us a peak into what European Big Tech would look like. And it demonstrates why building competitors to the commercial platforms of the US is doomed to end up just like it but with a European flag on top of its headquarters.
Thanks to Astrid Christofori and Lia Belle for pointing me to helpful info on atproto. Header photo by Jon Tyson on Unsplash